Summary On the morning of 22 May 2006, while at anchor in the Bay of Sept-les, an abandon-ship drill was carried out on board the bulk carrier SeaUrchin. At 1150eastern daylight time, during the recovery of the starboard lifeboat, the aft hook-release gear mechanism opened. The forward mechanism, unable to take the full load, also opened. The lifeboat then fell 11m, stern first, into the sea, fatally injuring one of the five crew members on board. Ce rapport est galement disponible en franais. Other Factual Information Particulars of the Vessels Description of the Vessel Photo 1 - SeaUrchin in Bay of Sept-leson the St. Lawrence River The SeaUrchin is a Panamax gearless bulk carrier with seven cargo holds. The accommodation, bridge, and engine room are located aft (see Photo1). The vessel is equipped with two identical lifeboats, located outboard of the accommodation, one on each side. Description of the Lifeboats Photo 2 - Port lifeboat alongside theSeaUrchin There are two davit-launched gravity-type enclosed lifeboats (model SZ-53BR). The lifeboats are self-propelled, self-righting, and fitted with hook-release gear mechanisms. They were manufactured in April2001 in Osaka, Japan (see Photo2). This lifeboat type was approved by the vessel's classification society, Nippon Kaiji Kyokai (ClassNK), under the approval number N-478. This design uses a remote-launching system intended to allow 25persons to board the lifeboat before leaving the vessel. History of the Voyage The SeaUrchin departed Bunbury, Australia, on 15April2006, with a cargo of alumina for Aluminerie Alouette at Sept-les, Quebec. The vessel arrived at Sept-les on 22May2006 and anchored in the bay at 0800,2 2.5nautical miles north of the Aluminerie Alouette plant at the approximate latitude and longitude of 5010'02"N, 06629'00"W (see AppendixA). That same day, the chief officer (C/O) carried out a lifeboat drill to familiarize and train crew members with the lifeboats. For the majority of participating crew members, this was their first practical abandon-ship drill on that vessel since joining. At approximately 1020, the C/O requested that the second officer (2/O), third officer (3/O), as well as the boatswain, two ordinary seamen (OS), three able seamen (AB), and the electrician, muster near the port lifeboat for a pre-drill training session. Once crew members donned their lifejackets, they boarded the lifeboat. While the C/O stood outside on deck, he gave step-by-step instructions on how to release the painter and activate the hook-release gear mechanism. After disembarking, crew members were asked to confirm if they understood the operation; they responded in the affirmative. Two drills were to be carried out subsequently. Five crew members, including the 2/O, and the 3/O- who was in charge of the port lifeboat- re-boarded for the first drill. Once the five crew members had fastened their seat belts, the C/O ordered the lifeboat to be lowered. When the lifeboat entered the water, the C/O ordered the hooks released and thereafter for the hook-release gear to be reset. To this end, two crew members inside the craft pulled the hooks inward to the reset position; another crew member pushed the release handle down in place, the safety pin was inserted, and the davit fall long links set in the hooks. When all aboard returned to their seats and were belted, the C/O ordered the lifeboat raised. The hoisting of the lifeboat was uneventful. Afterward, the crew disembarked and proceeded to the starboard lifeboat for the second drill. The 2/O, in charge of the starboard lifeboat, felt uncomfortable as it was his first time on the SeaUrchin and asked the 3/O to join him. With them were two ABs and anOS. The C/O and other crew members remained on deck. As with the first drill, the starboard lifeboat was lowered and, when it was afloat, the C/O ordered the hooks released. Once done, he ordered the hook-release gear to be reset. The 3/O reportedly had difficulty inserting the safety pin, but accomplished this with the help of another crew member. Eventually, he reported that everything was in place with the davit fall long links set in the hooks; the C/O subsequently ordered the lifeboat raised. At 1150, as the davit arms started to rock back in, the aft hook released. Subsequently, the forward hook released and the lifeboat fell 11m, hitting the water stern first at an approximate angle of 70. The craft then capsized on its port side and began taking on water. The master was immediately informed, and the general alarm sounded. At 1153, the master called the Marine Communications and Traffic Services (MCTS) centre to request assistance. The lifeboat ruptured upon hitting the water, creating an opening through which two crew members exited into the sea. One was subsequently rescued using a life ring and a rope ladder, and the other boarded the port lifeboat after it was lowered. Two other crew members were able to exit through the lifeboat's forward hatch and re-board the SeaUrchin. A head count, however, determined that the 2/O was missing, and the master informed the MCTS centre. In response, at 1226, an MCTS officer radioed a Mayday Relay, requesting assistance from nearby vessels. SeaUrchin crew members raised the forward end of the lifeboat using the provision crane. When the port access door surfaced, the body of the 2/O could be seen inside. Resuscitation efforts were attempted but failed. Launching and Recovery of Lifeboat Launch Procedure Launch can be initiated from inside the lifeboat- by pulling down the control wire linked to the davits' winch brake release mechanism- or by manually operating the brake from the vessel's boat deck. Recovery Procedure Once the davit fall long links are set in the hooks, the lifeboat is raised and, when contact is made between the floating block and the davit arms, the lifeboat is swung inward into its stowage position. Hook-Release Gear Mechanism There are a large number of often complex designs for release gear. The hook-release gear mechanism used in this occurrence, type SRS-37, is an on-load/off-load system. As shown in Figure1, it is composed of forward and aft hooks(A), a release handle(B), and release cables(C). A hydrostatic interlock unit(D) prevents the accidental release of the hooks when the lifeboat is not in the water. Figure1.Hook-release gear mechanism on the lifeboat The hook assembly is composed of three main components: the hook, the lock piece, and the cam (see Figure2). For simultaneous operation of the hooks, the release handle in way of the steering station is used. The release cables link the release handle to the cam lever of both hooks. Figure2.Hook assembly in locked position Resetting the release gear mechanism involves resetting the hooks and locking the mechanism Overall, this two-step task requires three people: one person at the steering station, and two other persons- that is, one at each hook. Step 1: Resetting the Hooks (Figures 3 to 6) The hook is pulled inward to adopt an upright position ready for hoisting. Note how the hook tail(B) fits into the notch of the lock piece (A), exerting a rotating force that lifts the lock piece (see Figure3) and allows for the full rotation of the cam(C) underneath (in step2). The design requires the user to apply a fair amount of force on the hook to lift the lock piece into final position (see Figure4). Figure3.Hook tail (B) lifting the lockpiece as it enters into thenotch (A) Figure4.Final position: the lock pieceis lifted and the hook tailinserted into the notch If the reset has been done properly, clearance (yellow lines) is visible between the cam and lock piece (see Figure 5). Note that this clearance must be maintained using constant pressure on the hook to keep the lock piece in place. If this clearance is not maintained, the lock piece will pivot back down and interfere with the cam, preventing the resetting of the cam into its final position (see Figure 6). Figure5.Yellow lines indicate clearancebetween lock piece and cam Figure6.Interference between lockpiece and cam Step 2: Locking the Mechanism (Figures 7 to 10) The second step requires pushing the release handle down (yellow arrow, F), thereby moving the release cables in the direction of the red arrow (see Figure7); the release cable will in turn push the lever, causing the cam to rotate (see Figure8). Note that, if the lock piece has pivoted down and interferes with the cam, the release handle will either not move or will only move following the exertion of an atypical amount of force. Moreover, if an undue amount of force is applied on the release handle, this force can cause deformation of the cable-attachment bracket and/or the release handle. Figure7.Release handle at steering station being pushed down Figure8.Cam being rotated into finalposition, with lock piece heldin place by force on hook Once the release handle has been pushed down to its final position, the holes in the release handle and the safety bracket align, allowing for the insertion of the safety pin. The safety pin is then rotated to its locking position (see Figure9). Figure9.Safety pin inserted in handle and bracket twisted to lockin place Figure10.Final position: leverhorizontal (red knob) andgreen reference marksaligned Note: If the operation is carried out as designed, in the final reset position, the cam lever will be horizontal and the green reference marks on the cam lever assembly and the bracket will be aligned (see Figure10). Lifeboat Inspection Pre-Occurrence At the end of March 2006, the vessel called at Hyundai Vinashin Shipyard in Khanh Hoa, Vietnam, for a Special Class survey. This included a lifesaving-appliance survey, which in turn featured an overhaul, inspection, and test of the lifeboat's hook-release mechanism as well as an examination of the davits and a dynamic testing of the winch brake. These tests and the examination were all deemed satisfactory. At the time of the survey, the fall wires were replaced. Post-Occurrence Photo 3 - The 25-person lifeboat, as received atthe TSB Engineering Laboratory A post-occurrence examination of the port lifeboat revealed that its hook-release gear mechanism had been improperly reset during the drill on the day of the occurrence. The starboard lifeboat was transported to the TSB Engineering Laboratory for examination and testing (see Photo3). Training The SeaUrchin had a crew of 23, none of whom were familiar with the type of hook-release gear mechanism on board the lifeboats- with the exception of the C/O, who had familiarized himself with the documentation two months before the occurrence. Onboard records indicate that crew members carried out an abandon-ship drill on 31March2006, before departing the shipyard. Accompanying comments refer to the exercise as "fair, not quite satisfactory." There was no written confirmation that lifeboats were lowered at that time. Two additional abandon-ship drills were conducted at sea, one on 05April2006 and the other on 06May2006; both were described as "satisfactory," although neither instance involved lowering the lifeboats. In both cases, crews took their place inside their designated lifeboat while the C/O explained the operation of the release gear. There was neither a formal evaluation of the crew's knowledge of the release gear mechanism nor any hands-on system usage- crew members were nonetheless told to become familiar with the system. At the end of those training sessions, crew members were asked if they understood; they responded in the affirmative. Training Material Copies of the instruction manual for the type SRS-37 hook-release gear mechanism, as well as the lifeboat's operating and maintenance manual, were available in the crew's mess room.3 Following the 05April2006 abandon-ship drill, the C/O requested that these be read by the crew. The instruction manual for the hook-release gear mechanism features coloured pictograms with Japanese text, as well as a translation in English, which was the vessel's working language. Coloured pictograms and warnings were also posted inside the lifeboats. According to the manual, the hook is reset when a crew member pulls it inward. The manual mentions the need to check for clearance and the alignment of the green reference marks, but there is no mention of the need to maintain constant pressure on the hook until the cam is fully rotated. Similarly Equipped Vessels According to the manufacturer, approximately 500lifeboats were fitted with the type SRS-37 release gear mechanism. However, as its products are often sold through third parties, the manufacturer has stated that there were difficulties in contacting/tracing current owners. Additionally, type SRS-37 has been out of production for several years; a new production design no longer requires crew members to manipulate the hook during reset.